Spaces:
Sleeping
Sleeping
Commit ·
71ad36b
1
Parent(s): ef07635
refactor
Browse files- model/trainer_state.json +1187 -0
model/trainer_state.json
ADDED
|
@@ -0,0 +1,1187 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 650,
|
| 3 |
+
"best_metric": 1.6722568273544312,
|
| 4 |
+
"best_model_checkpoint": "/content/drive/MyDrive/mistral_aggressive_training/checkpoint-650",
|
| 5 |
+
"epoch": 1.9475262368815591,
|
| 6 |
+
"eval_steps": 25,
|
| 7 |
+
"global_step": 650,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.014992503748125937,
|
| 14 |
+
"grad_norm": 1.3109337091445923,
|
| 15 |
+
"learning_rate": 2.4242424242424244e-06,
|
| 16 |
+
"loss": 2.1417,
|
| 17 |
+
"step": 5
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.029985007496251874,
|
| 21 |
+
"grad_norm": 1.2851405143737793,
|
| 22 |
+
"learning_rate": 5.4545454545454545e-06,
|
| 23 |
+
"loss": 2.0593,
|
| 24 |
+
"step": 10
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.044977511244377814,
|
| 28 |
+
"grad_norm": 1.3587734699249268,
|
| 29 |
+
"learning_rate": 8.484848484848486e-06,
|
| 30 |
+
"loss": 2.0723,
|
| 31 |
+
"step": 15
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.05997001499250375,
|
| 35 |
+
"grad_norm": 1.1362162828445435,
|
| 36 |
+
"learning_rate": 1.1515151515151517e-05,
|
| 37 |
+
"loss": 2.0299,
|
| 38 |
+
"step": 20
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.07496251874062969,
|
| 42 |
+
"grad_norm": 1.0861225128173828,
|
| 43 |
+
"learning_rate": 1.4545454545454546e-05,
|
| 44 |
+
"loss": 2.0103,
|
| 45 |
+
"step": 25
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.07496251874062969,
|
| 49 |
+
"eval_loss": 1.9985558986663818,
|
| 50 |
+
"eval_model_preparation_time": 0.0233,
|
| 51 |
+
"eval_runtime": 464.1818,
|
| 52 |
+
"eval_samples_per_second": 2.001,
|
| 53 |
+
"eval_steps_per_second": 0.502,
|
| 54 |
+
"step": 25
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 0.08995502248875563,
|
| 58 |
+
"grad_norm": 1.1834365129470825,
|
| 59 |
+
"learning_rate": 1.7575757575757576e-05,
|
| 60 |
+
"loss": 1.9597,
|
| 61 |
+
"step": 30
|
| 62 |
+
},
|
| 63 |
+
{
|
| 64 |
+
"epoch": 0.10494752623688156,
|
| 65 |
+
"grad_norm": 0.9937852025032043,
|
| 66 |
+
"learning_rate": 1.9968503937007875e-05,
|
| 67 |
+
"loss": 1.9421,
|
| 68 |
+
"step": 35
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.1199400299850075,
|
| 72 |
+
"grad_norm": 1.0617958307266235,
|
| 73 |
+
"learning_rate": 1.9811023622047244e-05,
|
| 74 |
+
"loss": 2.024,
|
| 75 |
+
"step": 40
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 0.13493253373313344,
|
| 79 |
+
"grad_norm": 1.0851763486862183,
|
| 80 |
+
"learning_rate": 1.9653543307086616e-05,
|
| 81 |
+
"loss": 1.8302,
|
| 82 |
+
"step": 45
|
| 83 |
+
},
|
| 84 |
+
{
|
| 85 |
+
"epoch": 0.14992503748125938,
|
| 86 |
+
"grad_norm": 1.1397876739501953,
|
| 87 |
+
"learning_rate": 1.9496062992125985e-05,
|
| 88 |
+
"loss": 1.9633,
|
| 89 |
+
"step": 50
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 0.14992503748125938,
|
| 93 |
+
"eval_loss": 1.9225448369979858,
|
| 94 |
+
"eval_model_preparation_time": 0.0233,
|
| 95 |
+
"eval_runtime": 463.9066,
|
| 96 |
+
"eval_samples_per_second": 2.003,
|
| 97 |
+
"eval_steps_per_second": 0.502,
|
| 98 |
+
"step": 50
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.16491754122938532,
|
| 102 |
+
"grad_norm": 1.0546423196792603,
|
| 103 |
+
"learning_rate": 1.9338582677165357e-05,
|
| 104 |
+
"loss": 1.8928,
|
| 105 |
+
"step": 55
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.17991004497751126,
|
| 109 |
+
"grad_norm": 1.0976288318634033,
|
| 110 |
+
"learning_rate": 1.9181102362204726e-05,
|
| 111 |
+
"loss": 1.8646,
|
| 112 |
+
"step": 60
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 0.19490254872563717,
|
| 116 |
+
"grad_norm": 1.0164967775344849,
|
| 117 |
+
"learning_rate": 1.9023622047244098e-05,
|
| 118 |
+
"loss": 1.9247,
|
| 119 |
+
"step": 65
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 0.2098950524737631,
|
| 123 |
+
"grad_norm": 1.0609030723571777,
|
| 124 |
+
"learning_rate": 1.8866141732283464e-05,
|
| 125 |
+
"loss": 1.8895,
|
| 126 |
+
"step": 70
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"epoch": 0.22488755622188905,
|
| 130 |
+
"grad_norm": 1.2414894104003906,
|
| 131 |
+
"learning_rate": 1.8708661417322836e-05,
|
| 132 |
+
"loss": 1.9753,
|
| 133 |
+
"step": 75
|
| 134 |
+
},
|
| 135 |
+
{
|
| 136 |
+
"epoch": 0.22488755622188905,
|
| 137 |
+
"eval_loss": 1.8853719234466553,
|
| 138 |
+
"eval_model_preparation_time": 0.0233,
|
| 139 |
+
"eval_runtime": 463.658,
|
| 140 |
+
"eval_samples_per_second": 2.004,
|
| 141 |
+
"eval_steps_per_second": 0.503,
|
| 142 |
+
"step": 75
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.239880059970015,
|
| 146 |
+
"grad_norm": 1.1238614320755005,
|
| 147 |
+
"learning_rate": 1.8551181102362205e-05,
|
| 148 |
+
"loss": 1.8656,
|
| 149 |
+
"step": 80
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.25487256371814093,
|
| 153 |
+
"grad_norm": 1.143813967704773,
|
| 154 |
+
"learning_rate": 1.8393700787401577e-05,
|
| 155 |
+
"loss": 1.8317,
|
| 156 |
+
"step": 85
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.2698650674662669,
|
| 160 |
+
"grad_norm": 1.2623740434646606,
|
| 161 |
+
"learning_rate": 1.8236220472440946e-05,
|
| 162 |
+
"loss": 1.8669,
|
| 163 |
+
"step": 90
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.2848575712143928,
|
| 167 |
+
"grad_norm": 1.04632568359375,
|
| 168 |
+
"learning_rate": 1.8078740157480318e-05,
|
| 169 |
+
"loss": 1.7412,
|
| 170 |
+
"step": 95
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.29985007496251875,
|
| 174 |
+
"grad_norm": 1.0752465724945068,
|
| 175 |
+
"learning_rate": 1.7921259842519687e-05,
|
| 176 |
+
"loss": 1.808,
|
| 177 |
+
"step": 100
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.29985007496251875,
|
| 181 |
+
"eval_loss": 1.8590933084487915,
|
| 182 |
+
"eval_model_preparation_time": 0.0233,
|
| 183 |
+
"eval_runtime": 463.9821,
|
| 184 |
+
"eval_samples_per_second": 2.002,
|
| 185 |
+
"eval_steps_per_second": 0.502,
|
| 186 |
+
"step": 100
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 0.3148425787106447,
|
| 190 |
+
"grad_norm": 1.1382629871368408,
|
| 191 |
+
"learning_rate": 1.776377952755906e-05,
|
| 192 |
+
"loss": 1.8629,
|
| 193 |
+
"step": 105
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 0.32983508245877063,
|
| 197 |
+
"grad_norm": 1.4844969511032104,
|
| 198 |
+
"learning_rate": 1.7606299212598424e-05,
|
| 199 |
+
"loss": 1.8546,
|
| 200 |
+
"step": 110
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.3448275862068966,
|
| 204 |
+
"grad_norm": 1.1989675760269165,
|
| 205 |
+
"learning_rate": 1.7448818897637796e-05,
|
| 206 |
+
"loss": 1.8497,
|
| 207 |
+
"step": 115
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 0.3598200899550225,
|
| 211 |
+
"grad_norm": 1.2271519899368286,
|
| 212 |
+
"learning_rate": 1.7291338582677165e-05,
|
| 213 |
+
"loss": 1.896,
|
| 214 |
+
"step": 120
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 0.3748125937031484,
|
| 218 |
+
"grad_norm": 1.179254412651062,
|
| 219 |
+
"learning_rate": 1.7133858267716537e-05,
|
| 220 |
+
"loss": 1.8161,
|
| 221 |
+
"step": 125
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 0.3748125937031484,
|
| 225 |
+
"eval_loss": 1.836364507675171,
|
| 226 |
+
"eval_model_preparation_time": 0.0233,
|
| 227 |
+
"eval_runtime": 464.1432,
|
| 228 |
+
"eval_samples_per_second": 2.002,
|
| 229 |
+
"eval_steps_per_second": 0.502,
|
| 230 |
+
"step": 125
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.38980509745127434,
|
| 234 |
+
"grad_norm": 1.3870253562927246,
|
| 235 |
+
"learning_rate": 1.6976377952755906e-05,
|
| 236 |
+
"loss": 1.8299,
|
| 237 |
+
"step": 130
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 0.4047976011994003,
|
| 241 |
+
"grad_norm": 1.1159974336624146,
|
| 242 |
+
"learning_rate": 1.6818897637795278e-05,
|
| 243 |
+
"loss": 1.7106,
|
| 244 |
+
"step": 135
|
| 245 |
+
},
|
| 246 |
+
{
|
| 247 |
+
"epoch": 0.4197901049475262,
|
| 248 |
+
"grad_norm": 1.2132537364959717,
|
| 249 |
+
"learning_rate": 1.6661417322834647e-05,
|
| 250 |
+
"loss": 1.8106,
|
| 251 |
+
"step": 140
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 0.43478260869565216,
|
| 255 |
+
"grad_norm": 1.180285930633545,
|
| 256 |
+
"learning_rate": 1.650393700787402e-05,
|
| 257 |
+
"loss": 1.7911,
|
| 258 |
+
"step": 145
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 0.4497751124437781,
|
| 262 |
+
"grad_norm": 1.2773538827896118,
|
| 263 |
+
"learning_rate": 1.6346456692913385e-05,
|
| 264 |
+
"loss": 1.8412,
|
| 265 |
+
"step": 150
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 0.4497751124437781,
|
| 269 |
+
"eval_loss": 1.817762017250061,
|
| 270 |
+
"eval_model_preparation_time": 0.0233,
|
| 271 |
+
"eval_runtime": 463.8218,
|
| 272 |
+
"eval_samples_per_second": 2.003,
|
| 273 |
+
"eval_steps_per_second": 0.502,
|
| 274 |
+
"step": 150
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 0.46476761619190404,
|
| 278 |
+
"grad_norm": 1.2192872762680054,
|
| 279 |
+
"learning_rate": 1.6188976377952757e-05,
|
| 280 |
+
"loss": 1.8606,
|
| 281 |
+
"step": 155
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 0.47976011994003,
|
| 285 |
+
"grad_norm": 1.4267574548721313,
|
| 286 |
+
"learning_rate": 1.6031496062992126e-05,
|
| 287 |
+
"loss": 1.8088,
|
| 288 |
+
"step": 160
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 0.4947526236881559,
|
| 292 |
+
"grad_norm": 1.4275727272033691,
|
| 293 |
+
"learning_rate": 1.5874015748031498e-05,
|
| 294 |
+
"loss": 1.7979,
|
| 295 |
+
"step": 165
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"epoch": 0.5097451274362819,
|
| 299 |
+
"grad_norm": 1.3130961656570435,
|
| 300 |
+
"learning_rate": 1.5716535433070866e-05,
|
| 301 |
+
"loss": 1.8059,
|
| 302 |
+
"step": 170
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 0.5247376311844077,
|
| 306 |
+
"grad_norm": 1.2740048170089722,
|
| 307 |
+
"learning_rate": 1.555905511811024e-05,
|
| 308 |
+
"loss": 1.7689,
|
| 309 |
+
"step": 175
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 0.5247376311844077,
|
| 313 |
+
"eval_loss": 1.801321268081665,
|
| 314 |
+
"eval_model_preparation_time": 0.0233,
|
| 315 |
+
"eval_runtime": 463.8247,
|
| 316 |
+
"eval_samples_per_second": 2.003,
|
| 317 |
+
"eval_steps_per_second": 0.502,
|
| 318 |
+
"step": 175
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.5397301349325337,
|
| 322 |
+
"grad_norm": 1.1887905597686768,
|
| 323 |
+
"learning_rate": 1.5401574803149607e-05,
|
| 324 |
+
"loss": 1.7052,
|
| 325 |
+
"step": 180
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.5547226386806596,
|
| 329 |
+
"grad_norm": 1.3410052061080933,
|
| 330 |
+
"learning_rate": 1.5244094488188978e-05,
|
| 331 |
+
"loss": 1.7916,
|
| 332 |
+
"step": 185
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.5697151424287856,
|
| 336 |
+
"grad_norm": 1.2976702451705933,
|
| 337 |
+
"learning_rate": 1.5086614173228347e-05,
|
| 338 |
+
"loss": 1.6751,
|
| 339 |
+
"step": 190
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.5847076461769115,
|
| 343 |
+
"grad_norm": 1.3378161191940308,
|
| 344 |
+
"learning_rate": 1.4929133858267717e-05,
|
| 345 |
+
"loss": 1.6846,
|
| 346 |
+
"step": 195
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.5997001499250375,
|
| 350 |
+
"grad_norm": 1.3700908422470093,
|
| 351 |
+
"learning_rate": 1.4771653543307088e-05,
|
| 352 |
+
"loss": 1.7537,
|
| 353 |
+
"step": 200
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.5997001499250375,
|
| 357 |
+
"eval_loss": 1.786393165588379,
|
| 358 |
+
"eval_model_preparation_time": 0.0233,
|
| 359 |
+
"eval_runtime": 464.0429,
|
| 360 |
+
"eval_samples_per_second": 2.002,
|
| 361 |
+
"eval_steps_per_second": 0.502,
|
| 362 |
+
"step": 200
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 0.6146926536731634,
|
| 366 |
+
"grad_norm": 1.3000359535217285,
|
| 367 |
+
"learning_rate": 1.4614173228346458e-05,
|
| 368 |
+
"loss": 1.7362,
|
| 369 |
+
"step": 205
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 0.6296851574212894,
|
| 373 |
+
"grad_norm": 1.3432403802871704,
|
| 374 |
+
"learning_rate": 1.4456692913385829e-05,
|
| 375 |
+
"loss": 1.7752,
|
| 376 |
+
"step": 210
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"epoch": 0.6446776611694153,
|
| 380 |
+
"grad_norm": 1.4008256196975708,
|
| 381 |
+
"learning_rate": 1.4299212598425199e-05,
|
| 382 |
+
"loss": 1.7294,
|
| 383 |
+
"step": 215
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 0.6596701649175413,
|
| 387 |
+
"grad_norm": 1.2891970872879028,
|
| 388 |
+
"learning_rate": 1.4141732283464568e-05,
|
| 389 |
+
"loss": 1.7085,
|
| 390 |
+
"step": 220
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"epoch": 0.6746626686656672,
|
| 394 |
+
"grad_norm": 1.1324400901794434,
|
| 395 |
+
"learning_rate": 1.3984251968503938e-05,
|
| 396 |
+
"loss": 1.7697,
|
| 397 |
+
"step": 225
|
| 398 |
+
},
|
| 399 |
+
{
|
| 400 |
+
"epoch": 0.6746626686656672,
|
| 401 |
+
"eval_loss": 1.7725938558578491,
|
| 402 |
+
"eval_model_preparation_time": 0.0233,
|
| 403 |
+
"eval_runtime": 464.0819,
|
| 404 |
+
"eval_samples_per_second": 2.002,
|
| 405 |
+
"eval_steps_per_second": 0.502,
|
| 406 |
+
"step": 225
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 0.6896551724137931,
|
| 410 |
+
"grad_norm": 1.4434219598770142,
|
| 411 |
+
"learning_rate": 1.3826771653543307e-05,
|
| 412 |
+
"loss": 1.7308,
|
| 413 |
+
"step": 230
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 0.704647676161919,
|
| 417 |
+
"grad_norm": 1.3460474014282227,
|
| 418 |
+
"learning_rate": 1.3669291338582678e-05,
|
| 419 |
+
"loss": 1.7323,
|
| 420 |
+
"step": 235
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 0.719640179910045,
|
| 424 |
+
"grad_norm": 1.2641957998275757,
|
| 425 |
+
"learning_rate": 1.3511811023622048e-05,
|
| 426 |
+
"loss": 1.6987,
|
| 427 |
+
"step": 240
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"epoch": 0.7346326836581709,
|
| 431 |
+
"grad_norm": 1.3505923748016357,
|
| 432 |
+
"learning_rate": 1.3354330708661419e-05,
|
| 433 |
+
"loss": 1.6961,
|
| 434 |
+
"step": 245
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.7496251874062968,
|
| 438 |
+
"grad_norm": 1.4946131706237793,
|
| 439 |
+
"learning_rate": 1.3196850393700789e-05,
|
| 440 |
+
"loss": 1.733,
|
| 441 |
+
"step": 250
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 0.7496251874062968,
|
| 445 |
+
"eval_loss": 1.7611132860183716,
|
| 446 |
+
"eval_model_preparation_time": 0.0233,
|
| 447 |
+
"eval_runtime": 463.8197,
|
| 448 |
+
"eval_samples_per_second": 2.003,
|
| 449 |
+
"eval_steps_per_second": 0.502,
|
| 450 |
+
"step": 250
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.7646176911544228,
|
| 454 |
+
"grad_norm": 1.433475375175476,
|
| 455 |
+
"learning_rate": 1.303937007874016e-05,
|
| 456 |
+
"loss": 1.8026,
|
| 457 |
+
"step": 255
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.7796101949025487,
|
| 461 |
+
"grad_norm": 1.448909044265747,
|
| 462 |
+
"learning_rate": 1.288188976377953e-05,
|
| 463 |
+
"loss": 1.7112,
|
| 464 |
+
"step": 260
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.7946026986506747,
|
| 468 |
+
"grad_norm": 1.3357568979263306,
|
| 469 |
+
"learning_rate": 1.2724409448818899e-05,
|
| 470 |
+
"loss": 1.6413,
|
| 471 |
+
"step": 265
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.8095952023988006,
|
| 475 |
+
"grad_norm": 1.4122332334518433,
|
| 476 |
+
"learning_rate": 1.2566929133858268e-05,
|
| 477 |
+
"loss": 1.7038,
|
| 478 |
+
"step": 270
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.8245877061469266,
|
| 482 |
+
"grad_norm": 1.5254954099655151,
|
| 483 |
+
"learning_rate": 1.2409448818897638e-05,
|
| 484 |
+
"loss": 1.6762,
|
| 485 |
+
"step": 275
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.8245877061469266,
|
| 489 |
+
"eval_loss": 1.7509944438934326,
|
| 490 |
+
"eval_model_preparation_time": 0.0233,
|
| 491 |
+
"eval_runtime": 464.0835,
|
| 492 |
+
"eval_samples_per_second": 2.002,
|
| 493 |
+
"eval_steps_per_second": 0.502,
|
| 494 |
+
"step": 275
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 0.8395802098950524,
|
| 498 |
+
"grad_norm": 1.4556777477264404,
|
| 499 |
+
"learning_rate": 1.2251968503937009e-05,
|
| 500 |
+
"loss": 1.7421,
|
| 501 |
+
"step": 280
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 0.8545727136431784,
|
| 505 |
+
"grad_norm": 1.2593179941177368,
|
| 506 |
+
"learning_rate": 1.2094488188976379e-05,
|
| 507 |
+
"loss": 1.7368,
|
| 508 |
+
"step": 285
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 0.8695652173913043,
|
| 512 |
+
"grad_norm": 1.5513230562210083,
|
| 513 |
+
"learning_rate": 1.193700787401575e-05,
|
| 514 |
+
"loss": 1.6937,
|
| 515 |
+
"step": 290
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 0.8845577211394303,
|
| 519 |
+
"grad_norm": 1.450356125831604,
|
| 520 |
+
"learning_rate": 1.177952755905512e-05,
|
| 521 |
+
"loss": 1.6819,
|
| 522 |
+
"step": 295
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 0.8995502248875562,
|
| 526 |
+
"grad_norm": 1.4953676462173462,
|
| 527 |
+
"learning_rate": 1.162204724409449e-05,
|
| 528 |
+
"loss": 1.7522,
|
| 529 |
+
"step": 300
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 0.8995502248875562,
|
| 533 |
+
"eval_loss": 1.7417070865631104,
|
| 534 |
+
"eval_model_preparation_time": 0.0233,
|
| 535 |
+
"eval_runtime": 463.9637,
|
| 536 |
+
"eval_samples_per_second": 2.002,
|
| 537 |
+
"eval_steps_per_second": 0.502,
|
| 538 |
+
"step": 300
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 0.9145427286356822,
|
| 542 |
+
"grad_norm": 1.3364105224609375,
|
| 543 |
+
"learning_rate": 1.1464566929133861e-05,
|
| 544 |
+
"loss": 1.7016,
|
| 545 |
+
"step": 305
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 0.9295352323838081,
|
| 549 |
+
"grad_norm": 1.7076566219329834,
|
| 550 |
+
"learning_rate": 1.1307086614173228e-05,
|
| 551 |
+
"loss": 1.7504,
|
| 552 |
+
"step": 310
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 0.9445277361319341,
|
| 556 |
+
"grad_norm": 1.397580862045288,
|
| 557 |
+
"learning_rate": 1.1149606299212599e-05,
|
| 558 |
+
"loss": 1.6943,
|
| 559 |
+
"step": 315
|
| 560 |
+
},
|
| 561 |
+
{
|
| 562 |
+
"epoch": 0.95952023988006,
|
| 563 |
+
"grad_norm": 1.4239177703857422,
|
| 564 |
+
"learning_rate": 1.0992125984251969e-05,
|
| 565 |
+
"loss": 1.6791,
|
| 566 |
+
"step": 320
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 0.974512743628186,
|
| 570 |
+
"grad_norm": 1.4564177989959717,
|
| 571 |
+
"learning_rate": 1.083464566929134e-05,
|
| 572 |
+
"loss": 1.7237,
|
| 573 |
+
"step": 325
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"epoch": 0.974512743628186,
|
| 577 |
+
"eval_loss": 1.7306653261184692,
|
| 578 |
+
"eval_model_preparation_time": 0.0233,
|
| 579 |
+
"eval_runtime": 464.0457,
|
| 580 |
+
"eval_samples_per_second": 2.002,
|
| 581 |
+
"eval_steps_per_second": 0.502,
|
| 582 |
+
"step": 325
|
| 583 |
+
},
|
| 584 |
+
{
|
| 585 |
+
"epoch": 0.9895052473763118,
|
| 586 |
+
"grad_norm": 1.4340523481369019,
|
| 587 |
+
"learning_rate": 1.067716535433071e-05,
|
| 588 |
+
"loss": 1.7005,
|
| 589 |
+
"step": 330
|
| 590 |
+
},
|
| 591 |
+
{
|
| 592 |
+
"epoch": 1.0029985007496252,
|
| 593 |
+
"grad_norm": 1.3697952032089233,
|
| 594 |
+
"learning_rate": 1.051968503937008e-05,
|
| 595 |
+
"loss": 1.6647,
|
| 596 |
+
"step": 335
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 1.0179910044977512,
|
| 600 |
+
"grad_norm": 1.4047369956970215,
|
| 601 |
+
"learning_rate": 1.0362204724409451e-05,
|
| 602 |
+
"loss": 1.7556,
|
| 603 |
+
"step": 340
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 1.0329835082458771,
|
| 607 |
+
"grad_norm": 1.493369460105896,
|
| 608 |
+
"learning_rate": 1.0204724409448821e-05,
|
| 609 |
+
"loss": 1.6457,
|
| 610 |
+
"step": 345
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"epoch": 1.047976011994003,
|
| 614 |
+
"grad_norm": 1.6234937906265259,
|
| 615 |
+
"learning_rate": 1.0047244094488188e-05,
|
| 616 |
+
"loss": 1.6274,
|
| 617 |
+
"step": 350
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 1.047976011994003,
|
| 621 |
+
"eval_loss": 1.7234022617340088,
|
| 622 |
+
"eval_model_preparation_time": 0.0233,
|
| 623 |
+
"eval_runtime": 463.754,
|
| 624 |
+
"eval_samples_per_second": 2.003,
|
| 625 |
+
"eval_steps_per_second": 0.502,
|
| 626 |
+
"step": 350
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.062968515742129,
|
| 630 |
+
"grad_norm": 1.6586315631866455,
|
| 631 |
+
"learning_rate": 9.88976377952756e-06,
|
| 632 |
+
"loss": 1.6198,
|
| 633 |
+
"step": 355
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 1.077961019490255,
|
| 637 |
+
"grad_norm": 1.5955413579940796,
|
| 638 |
+
"learning_rate": 9.73228346456693e-06,
|
| 639 |
+
"loss": 1.5932,
|
| 640 |
+
"step": 360
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 1.092953523238381,
|
| 644 |
+
"grad_norm": 1.6776100397109985,
|
| 645 |
+
"learning_rate": 9.5748031496063e-06,
|
| 646 |
+
"loss": 1.6465,
|
| 647 |
+
"step": 365
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 1.1079460269865067,
|
| 651 |
+
"grad_norm": 1.4672406911849976,
|
| 652 |
+
"learning_rate": 9.41732283464567e-06,
|
| 653 |
+
"loss": 1.6208,
|
| 654 |
+
"step": 370
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 1.1229385307346327,
|
| 658 |
+
"grad_norm": 1.507462978363037,
|
| 659 |
+
"learning_rate": 9.259842519685041e-06,
|
| 660 |
+
"loss": 1.6205,
|
| 661 |
+
"step": 375
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 1.1229385307346327,
|
| 665 |
+
"eval_loss": 1.7159619331359863,
|
| 666 |
+
"eval_model_preparation_time": 0.0233,
|
| 667 |
+
"eval_runtime": 463.9627,
|
| 668 |
+
"eval_samples_per_second": 2.002,
|
| 669 |
+
"eval_steps_per_second": 0.502,
|
| 670 |
+
"step": 375
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"epoch": 1.1379310344827587,
|
| 674 |
+
"grad_norm": 1.6919342279434204,
|
| 675 |
+
"learning_rate": 9.10236220472441e-06,
|
| 676 |
+
"loss": 1.6921,
|
| 677 |
+
"step": 380
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 1.1529235382308847,
|
| 681 |
+
"grad_norm": 1.569090723991394,
|
| 682 |
+
"learning_rate": 8.94488188976378e-06,
|
| 683 |
+
"loss": 1.6457,
|
| 684 |
+
"step": 385
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 1.1679160419790104,
|
| 688 |
+
"grad_norm": 1.6199414730072021,
|
| 689 |
+
"learning_rate": 8.78740157480315e-06,
|
| 690 |
+
"loss": 1.6988,
|
| 691 |
+
"step": 390
|
| 692 |
+
},
|
| 693 |
+
{
|
| 694 |
+
"epoch": 1.1829085457271364,
|
| 695 |
+
"grad_norm": 1.4945182800292969,
|
| 696 |
+
"learning_rate": 8.629921259842521e-06,
|
| 697 |
+
"loss": 1.5946,
|
| 698 |
+
"step": 395
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 1.1979010494752624,
|
| 702 |
+
"grad_norm": 1.6097604036331177,
|
| 703 |
+
"learning_rate": 8.47244094488189e-06,
|
| 704 |
+
"loss": 1.6567,
|
| 705 |
+
"step": 400
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 1.1979010494752624,
|
| 709 |
+
"eval_loss": 1.7092427015304565,
|
| 710 |
+
"eval_model_preparation_time": 0.0233,
|
| 711 |
+
"eval_runtime": 463.9844,
|
| 712 |
+
"eval_samples_per_second": 2.002,
|
| 713 |
+
"eval_steps_per_second": 0.502,
|
| 714 |
+
"step": 400
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 1.2128935532233882,
|
| 718 |
+
"grad_norm": 1.5406345129013062,
|
| 719 |
+
"learning_rate": 8.31496062992126e-06,
|
| 720 |
+
"loss": 1.5858,
|
| 721 |
+
"step": 405
|
| 722 |
+
},
|
| 723 |
+
{
|
| 724 |
+
"epoch": 1.2278860569715142,
|
| 725 |
+
"grad_norm": 1.7372868061065674,
|
| 726 |
+
"learning_rate": 8.157480314960631e-06,
|
| 727 |
+
"loss": 1.5864,
|
| 728 |
+
"step": 410
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 1.2428785607196402,
|
| 732 |
+
"grad_norm": 1.7523705959320068,
|
| 733 |
+
"learning_rate": 8.000000000000001e-06,
|
| 734 |
+
"loss": 1.6189,
|
| 735 |
+
"step": 415
|
| 736 |
+
},
|
| 737 |
+
{
|
| 738 |
+
"epoch": 1.2578710644677662,
|
| 739 |
+
"grad_norm": 1.6225429773330688,
|
| 740 |
+
"learning_rate": 7.84251968503937e-06,
|
| 741 |
+
"loss": 1.5691,
|
| 742 |
+
"step": 420
|
| 743 |
+
},
|
| 744 |
+
{
|
| 745 |
+
"epoch": 1.272863568215892,
|
| 746 |
+
"grad_norm": 1.6807804107666016,
|
| 747 |
+
"learning_rate": 7.68503937007874e-06,
|
| 748 |
+
"loss": 1.5915,
|
| 749 |
+
"step": 425
|
| 750 |
+
},
|
| 751 |
+
{
|
| 752 |
+
"epoch": 1.272863568215892,
|
| 753 |
+
"eval_loss": 1.7035413980484009,
|
| 754 |
+
"eval_model_preparation_time": 0.0233,
|
| 755 |
+
"eval_runtime": 463.9113,
|
| 756 |
+
"eval_samples_per_second": 2.003,
|
| 757 |
+
"eval_steps_per_second": 0.502,
|
| 758 |
+
"step": 425
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 1.287856071964018,
|
| 762 |
+
"grad_norm": 1.6341338157653809,
|
| 763 |
+
"learning_rate": 7.527559055118111e-06,
|
| 764 |
+
"loss": 1.5694,
|
| 765 |
+
"step": 430
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 1.302848575712144,
|
| 769 |
+
"grad_norm": 1.6684147119522095,
|
| 770 |
+
"learning_rate": 7.3700787401574816e-06,
|
| 771 |
+
"loss": 1.7089,
|
| 772 |
+
"step": 435
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 1.31784107946027,
|
| 776 |
+
"grad_norm": 1.7899603843688965,
|
| 777 |
+
"learning_rate": 7.21259842519685e-06,
|
| 778 |
+
"loss": 1.613,
|
| 779 |
+
"step": 440
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 1.3328335832083957,
|
| 783 |
+
"grad_norm": 1.667357325553894,
|
| 784 |
+
"learning_rate": 7.055118110236221e-06,
|
| 785 |
+
"loss": 1.6017,
|
| 786 |
+
"step": 445
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 1.3478260869565217,
|
| 790 |
+
"grad_norm": 1.6792216300964355,
|
| 791 |
+
"learning_rate": 6.897637795275591e-06,
|
| 792 |
+
"loss": 1.6203,
|
| 793 |
+
"step": 450
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 1.3478260869565217,
|
| 797 |
+
"eval_loss": 1.6988191604614258,
|
| 798 |
+
"eval_model_preparation_time": 0.0233,
|
| 799 |
+
"eval_runtime": 463.913,
|
| 800 |
+
"eval_samples_per_second": 2.003,
|
| 801 |
+
"eval_steps_per_second": 0.502,
|
| 802 |
+
"step": 450
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 1.3628185907046477,
|
| 806 |
+
"grad_norm": 1.6012407541275024,
|
| 807 |
+
"learning_rate": 6.740157480314962e-06,
|
| 808 |
+
"loss": 1.6082,
|
| 809 |
+
"step": 455
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 1.3778110944527735,
|
| 813 |
+
"grad_norm": 1.7225656509399414,
|
| 814 |
+
"learning_rate": 6.5826771653543306e-06,
|
| 815 |
+
"loss": 1.5841,
|
| 816 |
+
"step": 460
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 1.3928035982008995,
|
| 820 |
+
"grad_norm": 1.6564732789993286,
|
| 821 |
+
"learning_rate": 6.425196850393701e-06,
|
| 822 |
+
"loss": 1.7057,
|
| 823 |
+
"step": 465
|
| 824 |
+
},
|
| 825 |
+
{
|
| 826 |
+
"epoch": 1.4077961019490255,
|
| 827 |
+
"grad_norm": 1.5676624774932861,
|
| 828 |
+
"learning_rate": 6.2677165354330715e-06,
|
| 829 |
+
"loss": 1.5861,
|
| 830 |
+
"step": 470
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 1.4227886056971515,
|
| 834 |
+
"grad_norm": 1.7391592264175415,
|
| 835 |
+
"learning_rate": 6.110236220472442e-06,
|
| 836 |
+
"loss": 1.6339,
|
| 837 |
+
"step": 475
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 1.4227886056971515,
|
| 841 |
+
"eval_loss": 1.6934857368469238,
|
| 842 |
+
"eval_model_preparation_time": 0.0233,
|
| 843 |
+
"eval_runtime": 463.8519,
|
| 844 |
+
"eval_samples_per_second": 2.003,
|
| 845 |
+
"eval_steps_per_second": 0.502,
|
| 846 |
+
"step": 475
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"epoch": 1.4377811094452775,
|
| 850 |
+
"grad_norm": 1.8171783685684204,
|
| 851 |
+
"learning_rate": 5.952755905511811e-06,
|
| 852 |
+
"loss": 1.5567,
|
| 853 |
+
"step": 480
|
| 854 |
+
},
|
| 855 |
+
{
|
| 856 |
+
"epoch": 1.4527736131934033,
|
| 857 |
+
"grad_norm": 1.6563059091567993,
|
| 858 |
+
"learning_rate": 5.795275590551181e-06,
|
| 859 |
+
"loss": 1.5669,
|
| 860 |
+
"step": 485
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 1.4677661169415293,
|
| 864 |
+
"grad_norm": 1.7603881359100342,
|
| 865 |
+
"learning_rate": 5.637795275590552e-06,
|
| 866 |
+
"loss": 1.6006,
|
| 867 |
+
"step": 490
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 1.4827586206896552,
|
| 871 |
+
"grad_norm": 1.951175570487976,
|
| 872 |
+
"learning_rate": 5.480314960629922e-06,
|
| 873 |
+
"loss": 1.7085,
|
| 874 |
+
"step": 495
|
| 875 |
+
},
|
| 876 |
+
{
|
| 877 |
+
"epoch": 1.497751124437781,
|
| 878 |
+
"grad_norm": 1.6208112239837646,
|
| 879 |
+
"learning_rate": 5.322834645669291e-06,
|
| 880 |
+
"loss": 1.6304,
|
| 881 |
+
"step": 500
|
| 882 |
+
},
|
| 883 |
+
{
|
| 884 |
+
"epoch": 1.497751124437781,
|
| 885 |
+
"eval_loss": 1.6872224807739258,
|
| 886 |
+
"eval_model_preparation_time": 0.0233,
|
| 887 |
+
"eval_runtime": 463.9678,
|
| 888 |
+
"eval_samples_per_second": 2.002,
|
| 889 |
+
"eval_steps_per_second": 0.502,
|
| 890 |
+
"step": 500
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 1.512743628185907,
|
| 894 |
+
"grad_norm": 1.71792471408844,
|
| 895 |
+
"learning_rate": 5.1653543307086615e-06,
|
| 896 |
+
"loss": 1.6648,
|
| 897 |
+
"step": 505
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 1.527736131934033,
|
| 901 |
+
"grad_norm": 1.71132493019104,
|
| 902 |
+
"learning_rate": 5.007874015748032e-06,
|
| 903 |
+
"loss": 1.594,
|
| 904 |
+
"step": 510
|
| 905 |
+
},
|
| 906 |
+
{
|
| 907 |
+
"epoch": 1.5427286356821588,
|
| 908 |
+
"grad_norm": 1.6670012474060059,
|
| 909 |
+
"learning_rate": 4.850393700787402e-06,
|
| 910 |
+
"loss": 1.5709,
|
| 911 |
+
"step": 515
|
| 912 |
+
},
|
| 913 |
+
{
|
| 914 |
+
"epoch": 1.557721139430285,
|
| 915 |
+
"grad_norm": 1.6870834827423096,
|
| 916 |
+
"learning_rate": 4.692913385826772e-06,
|
| 917 |
+
"loss": 1.6262,
|
| 918 |
+
"step": 520
|
| 919 |
+
},
|
| 920 |
+
{
|
| 921 |
+
"epoch": 1.5727136431784108,
|
| 922 |
+
"grad_norm": 1.8740547895431519,
|
| 923 |
+
"learning_rate": 4.535433070866142e-06,
|
| 924 |
+
"loss": 1.7051,
|
| 925 |
+
"step": 525
|
| 926 |
+
},
|
| 927 |
+
{
|
| 928 |
+
"epoch": 1.5727136431784108,
|
| 929 |
+
"eval_loss": 1.682806372642517,
|
| 930 |
+
"eval_model_preparation_time": 0.0233,
|
| 931 |
+
"eval_runtime": 464.0592,
|
| 932 |
+
"eval_samples_per_second": 2.002,
|
| 933 |
+
"eval_steps_per_second": 0.502,
|
| 934 |
+
"step": 525
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 1.5877061469265368,
|
| 938 |
+
"grad_norm": 1.6121946573257446,
|
| 939 |
+
"learning_rate": 4.377952755905512e-06,
|
| 940 |
+
"loss": 1.577,
|
| 941 |
+
"step": 530
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 1.6026986506746628,
|
| 945 |
+
"grad_norm": 1.8216692209243774,
|
| 946 |
+
"learning_rate": 4.220472440944882e-06,
|
| 947 |
+
"loss": 1.5505,
|
| 948 |
+
"step": 535
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 1.6176911544227885,
|
| 952 |
+
"grad_norm": 1.7587759494781494,
|
| 953 |
+
"learning_rate": 4.062992125984252e-06,
|
| 954 |
+
"loss": 1.6189,
|
| 955 |
+
"step": 540
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 1.6326836581709145,
|
| 959 |
+
"grad_norm": 1.8402221202850342,
|
| 960 |
+
"learning_rate": 3.905511811023622e-06,
|
| 961 |
+
"loss": 1.6749,
|
| 962 |
+
"step": 545
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 1.6476761619190405,
|
| 966 |
+
"grad_norm": 1.6589854955673218,
|
| 967 |
+
"learning_rate": 3.748031496062993e-06,
|
| 968 |
+
"loss": 1.7241,
|
| 969 |
+
"step": 550
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 1.6476761619190405,
|
| 973 |
+
"eval_loss": 1.6803463697433472,
|
| 974 |
+
"eval_model_preparation_time": 0.0233,
|
| 975 |
+
"eval_runtime": 463.9675,
|
| 976 |
+
"eval_samples_per_second": 2.002,
|
| 977 |
+
"eval_steps_per_second": 0.502,
|
| 978 |
+
"step": 550
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 1.6626686656671663,
|
| 982 |
+
"grad_norm": 1.6675294637680054,
|
| 983 |
+
"learning_rate": 3.5905511811023625e-06,
|
| 984 |
+
"loss": 1.5372,
|
| 985 |
+
"step": 555
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 1.6776611694152923,
|
| 989 |
+
"grad_norm": 1.6088923215866089,
|
| 990 |
+
"learning_rate": 3.433070866141733e-06,
|
| 991 |
+
"loss": 1.6013,
|
| 992 |
+
"step": 560
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 1.6926536731634183,
|
| 996 |
+
"grad_norm": 1.9371333122253418,
|
| 997 |
+
"learning_rate": 3.2755905511811026e-06,
|
| 998 |
+
"loss": 1.7105,
|
| 999 |
+
"step": 565
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 1.707646176911544,
|
| 1003 |
+
"grad_norm": 1.9125174283981323,
|
| 1004 |
+
"learning_rate": 3.118110236220473e-06,
|
| 1005 |
+
"loss": 1.6931,
|
| 1006 |
+
"step": 570
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"epoch": 1.7226386806596703,
|
| 1010 |
+
"grad_norm": 1.805245041847229,
|
| 1011 |
+
"learning_rate": 2.9606299212598427e-06,
|
| 1012 |
+
"loss": 1.5554,
|
| 1013 |
+
"step": 575
|
| 1014 |
+
},
|
| 1015 |
+
{
|
| 1016 |
+
"epoch": 1.7226386806596703,
|
| 1017 |
+
"eval_loss": 1.6775026321411133,
|
| 1018 |
+
"eval_model_preparation_time": 0.0233,
|
| 1019 |
+
"eval_runtime": 463.9093,
|
| 1020 |
+
"eval_samples_per_second": 2.003,
|
| 1021 |
+
"eval_steps_per_second": 0.502,
|
| 1022 |
+
"step": 575
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 1.737631184407796,
|
| 1026 |
+
"grad_norm": 1.753201961517334,
|
| 1027 |
+
"learning_rate": 2.803149606299213e-06,
|
| 1028 |
+
"loss": 1.5935,
|
| 1029 |
+
"step": 580
|
| 1030 |
+
},
|
| 1031 |
+
{
|
| 1032 |
+
"epoch": 1.752623688155922,
|
| 1033 |
+
"grad_norm": 1.9081984758377075,
|
| 1034 |
+
"learning_rate": 2.645669291338583e-06,
|
| 1035 |
+
"loss": 1.6316,
|
| 1036 |
+
"step": 585
|
| 1037 |
+
},
|
| 1038 |
+
{
|
| 1039 |
+
"epoch": 1.767616191904048,
|
| 1040 |
+
"grad_norm": 1.7977555990219116,
|
| 1041 |
+
"learning_rate": 2.488188976377953e-06,
|
| 1042 |
+
"loss": 1.5451,
|
| 1043 |
+
"step": 590
|
| 1044 |
+
},
|
| 1045 |
+
{
|
| 1046 |
+
"epoch": 1.7826086956521738,
|
| 1047 |
+
"grad_norm": 1.817696213722229,
|
| 1048 |
+
"learning_rate": 2.330708661417323e-06,
|
| 1049 |
+
"loss": 1.5799,
|
| 1050 |
+
"step": 595
|
| 1051 |
+
},
|
| 1052 |
+
{
|
| 1053 |
+
"epoch": 1.7976011994002998,
|
| 1054 |
+
"grad_norm": 1.7235620021820068,
|
| 1055 |
+
"learning_rate": 2.173228346456693e-06,
|
| 1056 |
+
"loss": 1.6196,
|
| 1057 |
+
"step": 600
|
| 1058 |
+
},
|
| 1059 |
+
{
|
| 1060 |
+
"epoch": 1.7976011994002998,
|
| 1061 |
+
"eval_loss": 1.6750439405441284,
|
| 1062 |
+
"eval_model_preparation_time": 0.0233,
|
| 1063 |
+
"eval_runtime": 464.07,
|
| 1064 |
+
"eval_samples_per_second": 2.002,
|
| 1065 |
+
"eval_steps_per_second": 0.502,
|
| 1066 |
+
"step": 600
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 1.8125937031484258,
|
| 1070 |
+
"grad_norm": 1.7125576734542847,
|
| 1071 |
+
"learning_rate": 2.015748031496063e-06,
|
| 1072 |
+
"loss": 1.5546,
|
| 1073 |
+
"step": 605
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 1.8275862068965516,
|
| 1077 |
+
"grad_norm": 1.693699598312378,
|
| 1078 |
+
"learning_rate": 1.8582677165354333e-06,
|
| 1079 |
+
"loss": 1.5906,
|
| 1080 |
+
"step": 610
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 1.8425787106446778,
|
| 1084 |
+
"grad_norm": 1.7908601760864258,
|
| 1085 |
+
"learning_rate": 1.7007874015748034e-06,
|
| 1086 |
+
"loss": 1.6616,
|
| 1087 |
+
"step": 615
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 1.8575712143928036,
|
| 1091 |
+
"grad_norm": 1.7427383661270142,
|
| 1092 |
+
"learning_rate": 1.5433070866141734e-06,
|
| 1093 |
+
"loss": 1.5276,
|
| 1094 |
+
"step": 620
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 1.8725637181409296,
|
| 1098 |
+
"grad_norm": 2.0088937282562256,
|
| 1099 |
+
"learning_rate": 1.3858267716535435e-06,
|
| 1100 |
+
"loss": 1.5377,
|
| 1101 |
+
"step": 625
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 1.8725637181409296,
|
| 1105 |
+
"eval_loss": 1.673450231552124,
|
| 1106 |
+
"eval_model_preparation_time": 0.0233,
|
| 1107 |
+
"eval_runtime": 464.0489,
|
| 1108 |
+
"eval_samples_per_second": 2.002,
|
| 1109 |
+
"eval_steps_per_second": 0.502,
|
| 1110 |
+
"step": 625
|
| 1111 |
+
},
|
| 1112 |
+
{
|
| 1113 |
+
"epoch": 1.8875562218890556,
|
| 1114 |
+
"grad_norm": 1.863299012184143,
|
| 1115 |
+
"learning_rate": 1.2283464566929135e-06,
|
| 1116 |
+
"loss": 1.545,
|
| 1117 |
+
"step": 630
|
| 1118 |
+
},
|
| 1119 |
+
{
|
| 1120 |
+
"epoch": 1.9025487256371814,
|
| 1121 |
+
"grad_norm": 1.5976277589797974,
|
| 1122 |
+
"learning_rate": 1.0708661417322836e-06,
|
| 1123 |
+
"loss": 1.6201,
|
| 1124 |
+
"step": 635
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 1.9175412293853074,
|
| 1128 |
+
"grad_norm": 1.758497953414917,
|
| 1129 |
+
"learning_rate": 9.133858267716536e-07,
|
| 1130 |
+
"loss": 1.6426,
|
| 1131 |
+
"step": 640
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"epoch": 1.9325337331334334,
|
| 1135 |
+
"grad_norm": 1.696065902709961,
|
| 1136 |
+
"learning_rate": 7.559055118110237e-07,
|
| 1137 |
+
"loss": 1.7097,
|
| 1138 |
+
"step": 645
|
| 1139 |
+
},
|
| 1140 |
+
{
|
| 1141 |
+
"epoch": 1.9475262368815591,
|
| 1142 |
+
"grad_norm": 1.9475734233856201,
|
| 1143 |
+
"learning_rate": 5.984251968503937e-07,
|
| 1144 |
+
"loss": 1.6048,
|
| 1145 |
+
"step": 650
|
| 1146 |
+
},
|
| 1147 |
+
{
|
| 1148 |
+
"epoch": 1.9475262368815591,
|
| 1149 |
+
"eval_loss": 1.6722568273544312,
|
| 1150 |
+
"eval_model_preparation_time": 0.0233,
|
| 1151 |
+
"eval_runtime": 463.9611,
|
| 1152 |
+
"eval_samples_per_second": 2.002,
|
| 1153 |
+
"eval_steps_per_second": 0.502,
|
| 1154 |
+
"step": 650
|
| 1155 |
+
}
|
| 1156 |
+
],
|
| 1157 |
+
"logging_steps": 5,
|
| 1158 |
+
"max_steps": 668,
|
| 1159 |
+
"num_input_tokens_seen": 0,
|
| 1160 |
+
"num_train_epochs": 2,
|
| 1161 |
+
"save_steps": 50,
|
| 1162 |
+
"stateful_callbacks": {
|
| 1163 |
+
"EarlyStoppingCallback": {
|
| 1164 |
+
"args": {
|
| 1165 |
+
"early_stopping_patience": 5,
|
| 1166 |
+
"early_stopping_threshold": 0.0005
|
| 1167 |
+
},
|
| 1168 |
+
"attributes": {
|
| 1169 |
+
"early_stopping_patience_counter": 0
|
| 1170 |
+
}
|
| 1171 |
+
},
|
| 1172 |
+
"TrainerControl": {
|
| 1173 |
+
"args": {
|
| 1174 |
+
"should_epoch_stop": false,
|
| 1175 |
+
"should_evaluate": false,
|
| 1176 |
+
"should_log": false,
|
| 1177 |
+
"should_save": true,
|
| 1178 |
+
"should_training_stop": false
|
| 1179 |
+
},
|
| 1180 |
+
"attributes": {}
|
| 1181 |
+
}
|
| 1182 |
+
},
|
| 1183 |
+
"total_flos": 9.115437053132145e+17,
|
| 1184 |
+
"train_batch_size": 4,
|
| 1185 |
+
"trial_name": null,
|
| 1186 |
+
"trial_params": null
|
| 1187 |
+
}
|